NDVI Images Research Articles

A New Winter Wheat Crop Segmentation Method Based on a New Fast-UNet Model and Multi-Temporal Sentinel-2 Images

Mapping and monitoring crops are the most complex and difficult tasks for experts processing and analyzing remote sensing (RS) images. Classifying crops using RS images is the most expensive task, and it requires intensive labor, especially in the sample collection phase. Fieldwork requires periodic visits to collect data about the crop’s physiochemical characteristics and separating them using the known conventional machine learning algorithms and remote sensing images. As the problem becomes more complex because of the diversity of crop types and the increase in area size, sample collection becomes more complex and unreliable. To avoid these problems, a new segmentation model was created that does not require sample collection or high-resolution images and can successfully distinguish wheat from other crops. Moreover, UNet is a well-known Convolutional Neural Network (CNN), and the semantic method was adjusted to become more powerful, faster, and use fewer resources. The new model was named Fast-UNet and was used to improve the segmentation of wheat crops. Fast-UNet was compared to UNet and Google’s newly developed semantic segmentation model, DeepLabV3+. The new model was faster than the compared models, and it had the highest average accuracy compared to UNet and DeepLabV3+, with values of 93.45, 93.05, and 92.56 respectively. Finally, new datasets of time series NDVI images and ground truth data were created. These datasets, and the newly developed model, were made available publicly on the Web.

Satellite-based soybean yield prediction in Argentina: A comparison between panel regression and deep learning methods

The accurate prediction of soybean yield is vital for global food market stabilization and food security. Recent advancements in remote sensing technology have significantly amplified interest in leveraging satellite-based methods for predicting crop yield. These methods offer in-season yield estimates. By utilizing this timely information, decision-makers can formulate strategic, well-informed choices that preemptively mitigate potential food price hikes, ultimately bolstering food security. While simple regression models have been widely utilized for satellite-based yield prediction, researchers have recently begun to explore the use of deep learning algorithms. This study compares the performance of panel regression and deep learning models for in-season soybean yield prediction at the Department (county-equivalent) level in Argentina. Data sources include the latest soybean land use products and MODIS bi-weekly vegetation index products. Results indicate that deep learning models significantly outperform panel regression. Deep learning Long Short-Term Memory (LSTM) models, which incorporate attention mechanism and a series of peak NDVI images, generate more accurate and time-sensitive predictions. Among competing LSTM models, the one with attention mechanism applied to the entire growing season's NDVI data yields the highest prediction accuracy, with a Root Mean Square Error (RMSE) of 505.78 kg/ha and Normalized Root Mean Square Error (NRMSE) of 0.0726. The LSTM model with attention on the three highest NDVI images attains a satisfactory prediction accuracy (RMSE = 627.28 kg/ha, NRMSE = 0.089) six weeks prior to harvest. This study presents a robust model for predicting crop yields, promoting sustainable production of soybeans and facilitating knowledgeable choices among farmers and policymakers.

Estimating high-density aboveground biomass within a complex tropical grassland using Worldview-3 imagery

A large percentage of native grassland ecosystems have been severely degraded as a result of urbanization and intensive commercial agriculture. Extensive nitrogen-based fertilization regimes are widely used to rehabilitate and boost productivity in these grasslands. As a result, modern management frameworks rely heavily on detailed and accurate information on vegetation condition to monitor the success of these interventions. However, in high-density environments, biomass signal saturation has hampered detailed monitoring of rangeland condition. This issue stems from traditional broad-band vegetation indices (such as NDVI) responding to high levels of photosynthetically active radiation (PAR) absorption by leaf chlorophyll, which affects leaf area index (LAI) sensitivity within densely vegetative regions. Whilst alternate hyperspectral solutions may alleviate the problem to a certain degree, they are often too costly and not readily available within developing regions. To this end, this study evaluated the use of high-resolution Worldview-3 imagery in combination with modified NDVI indices and image manipulation techniques in reducing the effects of biomass signal saturation within a complex tropical grassland. Using the random forest algorithm, several modified NDVI-type indices were developed from all potential dual-band combinations of the Worldview-3 image. Thereafter, linear contrast stretching and histogram equalization were implemented in conjunction with Singular Value Decomposition (SVD) to improve high-density biomass estimation. Results demonstrated that both contrast enhancement techniques, when combined with SVD, improved high-density biomass estimation. However, linear contrast stretching, SVD, and modified NDVI indices developed from the red (630–690 nm), green (510–580 nm), and near-infrared 1 (770–895 nm) bands were found to produce the best biomass predictive model (R2 = 0.71, RMSE = 0.40 kg/m2). The results generated from this research offer a means to alleviate the biomass saturation problem. This framework provides a platform to assist rangeland managers in regionally assessing changes in vegetation condition within high-density grasslands.

Decadal Variations in Area under Different Soil Erosion Classes using RUSLE and GIS: Case Studies of River Basins from Western and Eastern Arunachal Pradesh

Abstract This study provides a comparative evaluation of spatio-temporal distribution of soil erosion in Western (Mago Basin) and Eastern (Dibang Basin) basins of Arunachal Pradesh, India as these two basins are vulnerably exposed to soil erosion due to its topographical characteristics of mountainous steep slope and experiences heavy rainfall. The study was carried out for a ten-years period (2003 to 2014) using RUSLE model which encompasses five important factors contributing to soil erosion. Rainfall erosivity (R factor) map was calculated using Climate Prediction Center gridded precipitation. Soil map and soil samples were used to analyze soil erodibility (K factor) map. Slope length and slope steepness (LS factor) maps were computed from SRTM DEM (30 m resolution). MODIS NDVI images were used to obtain cover management (C factor) map. Landuse Landcover map was used to obtained support practice (P factor) map. Higher value in rainfall erosivity and cover management factor was observed in Mago basin which contributed to higher average annual soil loss of 17.423 t ha−1 y−1 in Mago basin and 5.461 t ha−1 y−1 in Dibang basin, whereas the other three factor values were almost the same. The spatial maps showed 56.65% of Mago basin area and 76.27% of Dibang basin area was under the class of slight erosion, with the remaining areas of moderate to severe erosion risk for both the basins. Temporal average soil erosion in Mago basin varied within moderate to very high erosion classes whereas Dibang basin erosion classes varied from slight to moderate. The temporal trend line showed that the overall soil erosion was increasing at an alarming rate for Mago basin whereas a slight increase in Dibang basin was observed.

Methods for estimating agricultural cropland yield based on the comparison of NDVI images analyzed by means of Image segmentation algorithms: A tool for spatial planning decisions

Methods for estimating agricultural cropland yield based on the comparison of NDVI images analyzed by means of Image segmentation algorithms: A tool for spatial planning decisions

Understanding the spatio-temporal behaviour of the sunflower crop for subfield areas delineation using Sentinel‐2 NDVI time-series images in an organic farming system

The study investigates the suitability of time series Sentinel-2 NDVI-derived maps for the subfield detection of a sunflower crop cultivated in an organic farming system. The aim was to understand the spatio-temporal behaviour of subfield areas identified by the K-means algorithm from NDVI maps obtained from satellite images and the ground yield data variability to increase the efficiency of delimiting management zones in an organic farming system.Experiments were conducted on a surface of 29 ha. NDVI time series derived from Sentinel-2 images and k-means algorithm for rapidly delineating the sunflower subfield areas were used. The crop achene yields in the whole field ranged from 1.3 to 3.77 t ha−1 with a significant within-field spatial variability. The cluster analysis of hand-sampled data showed three subfields with achene yield mean values of 3.54 t ha−1 (cluster 1), 2.98 t ha−1 (cluster 2), and 2.07 t ha−1 (Cluster 3). In the cluster analysis of NDVI data, the k-means algorithm has early delineated the subfield crop spatial and temporal yield variability. The best period for identifying subfield areas starts from the inflorescences development stage to the development of the fruit stage. Analyzing the NDVI subfield areas and yield data, it was found that cluster 1 covers an area of 42.4% of the total surface and 50% of the total achene yield; cluster 2 covers 35% of both surface and yield. Instead, the surface of cluster 3 covers 22.2% of the total surface with 15% of achene yield. K-means algorithm derived from Sentinel-2 NDVI images delineates the sunflower subfield areas.Sentinel-2 images and k-means algorithms can improve an efficient assessment of subfield areas in sunflower crops. Identifying subfield areas can lead to site-specific long-term agronomic actions for improving the sustainable intensification of agriculture in the organic farming system.

Analyzing the Characteristics of Cropping Intensity’s Change of Cultivated Land in China During 2010-2019

The sustained stability of the arable land replanting index is crucial to the national food security strategy. The exploration of the temporal and spatial changes in the arable land replanting index is of great significance for agricultural development and food security evaluation. This study investigates China’s arable land replanting index from 2010 to 2019 using MODIS NDVI image data, S-G filtering, and quadratic differentiation methods. The results show that there are significant spatial differences in China’s arable land replanting index, with the Huang-Huai-Hai region mainly producing double-cropping crops, the Northeast Plain and Loess Plateau mainly producing single-cropping crops, and the area south of the Yangtze River mainly producing multiple-cropping crops. Overall, China’s food production is mainly based on single-season crops. There is a gradual shift towards double-season crops from north to south, with lower replanting indices in the northwest and higher indices in the eastern provinces. The south has significantly higher indices than the north. During the study period, the arable land replanting index showed an overall upward trend. There were significant increases in the replanting index in the northeast, the Loess Plateau, and the northern Huang-Huai-Hai region. However, there was a downward trend in the southern Huang-Huai-Hai region and the middle and lower reaches of the Yangtze River. It is crucial to maintain the effective planting area of arable land in the Loess Plateau and the double-season planting area of arable land in the Huang-Huai-Hai region while also addressing the downward trend of the arable land replanting index in the middle and lower reaches of the Yangtze River to ensure food security by stabilizing the index.

Assessment of Burned Areas during the Pantanal Fire Crisis in 2020 Using Sentinel-2 Images

The Pantanal biome—a tropical wetland area—has been suffering a prolonged drought that started in 2019 and peaked in 2020. This favored the occurrence of natural disasters and led to the 2020 Pantanal fire crisis. The purpose of this work was to map the burned area’s extent during this crisis in the Brazilian portion of the Pantanal biome using Sentinel-2 MSI images. The classification of the burned areas was performed using a machine learning algorithm (Random Forest) in the Google Earth Engine platform. Input variables in the algorithm were the percentiles 10, 25, 50, 75, and 90 of monthly (July to December) mosaics of the shade fraction, NDVI, and NBR images derived from Sentinel-2 MSI images. The results showed an overall accuracy of 95.9% and an estimate of 44,998 km2 burned in the Brazilian portion of the Pantanal, which resulted in severe ecosystem destruction and biodiversity loss in this biome. The burned area estimated in this work was higher than those estimated by the MCD64A1 (35,837 km2), Fire_cci (36,017 km2), GABAM (14,307 km2), and MapBiomas Fogo (23,372 km2) burned area products, which presented lower accuracies. These differences can be explained by the distinct datasets and methods used to obtain those estimates. The proposed approach based on Sentinel-2 images can potentially refine the burned area’s estimation at a regional scale and, consequently, improve the estimate of trace gases and aerosols associated with biomass burning, where global biomass burning inventories are widely known for having biases at a regional scale. Our study brings to light the necessity of developing approaches that aim to improve data and theory about the impacts of fire in regions critically sensitive to climate change, such as the Pantanal, in order to improve Earth systems models that forecast wetland–atmosphere interactions, and the role of these fires on current and future climate change over these regions.

GIS AND REMOTE SENSINGBASED RECONNAISSANCE OF VEGETATION ANDAGROFORESTRY AREAS IN SOUTH 24 PARGANAS, WEST BENGAL, INDIA

Accessing the advantages of combining GIS with remote sensing could be better utilised for specifying vegetation cover including agroforestry systems in South 24 Parganas district with farm lands and ponds historically shaped out of mostly swampy lands covered with jungles and the Sundarbans crisscrossed with creeks and rivers in the south-easternmost part of West Bengal state in India by image processing based on updating of data or information in the form of maps or attributes. Such analysis for this district was performed mostly at the intervals of five years from 2005 to 2019.Changes in expanse of vegetation areas were classified in land coverage under deforestation and revegetation. Such changes are distinguishable through NDVI image values in the form of maps of the district and subdivisions, created on the Arc GIS 10.8 and ERDAS Imagine 2015 platforms by retrieving data of Landsat 4 - 5 TM Level 1 and Landsat 8 OLI/ TIRS Level 1 imageries as retrieved from USGS website. Healthy vegetation including agroforestry were sorted out through demarcation of highest indices of NDVI data. Areal extents of less than 10 ha of healthy vegetation were sorted out as areas either under new plantation or horti-agri or agroforestry. Individual agroforestry systems were found out with the help of Google Earth Pro. Monitoring of vegetation cover as well as agroforestry will be helpful to achieve SDG 15 and, thereby SDG 2 and 8 as defined by the UN for conservation of life on land as well as zero hunger including decent work and economic growth.

Study on Land Use/Land Cover and NDVI Changes in Alliveru Watershed Project of West Godavari District, Andhra Pradesh Using Remote Sensing and Geographical Information System

Government of India has taken up Integrated Watershed Development Program (IWMP), now Pradhan Mantri Krishi Sinchayee Yojana (PMKSY) watersheds to restore the ecological balance by harnessing, conserving, and developing degraded natural resources such as soil, vegetation, and water in rainfed areas of the country. In Alliveru Project of Buttaigudem Mandal, West Godavari district of Andhra Pradesh, the watershed programme was started during 20132014 and completed in 20202021. The present study was taken up to assess the impact of watershed activities on LULC and NDVI using Remote Sensing and GIS techniques. Terrain corrected Resourcesat-2 LISS IV for year 201415 and 20202,1Erdas Imagine 2015, and ArcGIS 10.8 software have been used in the present study. Onscreen visual interpretation technique has been used while carrying out LULC change analysis and image processing has been used to generate NDVI images. NDVI analysis has shown increase in the vegetative cover in the project area. Significant changes were observed in LULC over the project implementation period due to the soil and water conservation activities by the State Level Nodal Agency (SLNA) under Panchayat Raj & Rural Development (PR&RD) Department. The results showed reduction in wastelands by 21.10% and increase in cropland and plantation by 4.47% and 6.36% respectively. Another positive indicator of watershed development in the project area is the increase in waterbodies by 43.88% due to construction of rain water harvesting structures , that is, 43 check dams, 44 percolation tanks, 6 farm ponds and 2 mini percolation tanks.

Vegetation Classification in Urban Areas by Combining UAV-Based NDVI and Thermal Infrared Image

Vegetation has become very important decision-making information in promoting tasks such as urban regeneration, urban planning, environment, and landscaping. In the past, the vegetation index was calculated by combining images of various wavelength regions mainly acquired from the Landsat satellite’s TM or ETM+ sensor. Recently, a technology using UAV-based multispectral images has been developed to obtain more rapid and precise vegetation information. NDVI is a method of calculating the vegetation index by combining the red and near-infrared bands, and is currently the most widely used. In this study, NDVI was calculated using UAV-based multispectral images to classify vegetation. However, among the areas analyzed using NDVI, there was a problem that areas coated with urethane, such as basketball courts and waterproof coating roofs, were classified as vegetation areas. In order to examine these problems, the reflectance of each land cover was investigated using the ASD FieldSpec4 spectrometer. As a result of analyzing the spectrometer measurements, the NDVI values of basketball courts and waterproof coating roofs were similar to those of grass with slightly lower vegetation. To solve this problem, the temperature characteristics of the target site were analyzed using UAV-based thermal infrared images, and vegetation area was analyzed by combining the temperature information with NDVI. To evaluate the accuracy of the vegetation classification technology, 4409 verification points were selected, and kappa coefficients were analyzed for the method using only NDVI and the method using NDVI and thermal infrared images. Compared to the kappa coefficient of 0.830, which was analyzed by applying only NDVI, the kappa coefficient, which was analyzed by combining NDVI and thermal infrared images, was 0.934, which was higher. Therefore, it is very effective to apply a technology that classifies vegetation by combining NDVI and thermal infrared images in urban areas with many urethane-coated land cover such as basketball courts or waterproof coating roofs.

Assessment of Urban Green Space Distribution within the Scope of European Green Deal Using NDVI Indice; Case of Nicosia/Cyprus

In this study, the urban green space distribution was determined, and the city's compliance with the European Green Deal was evaluated. Firstly, Landsat 8 satellite images were radiometrically corrected with the Quantum GIS software. Then, NDVI was produced using the Red-Green-Blue and Near-Infrared bands of the processed data obtained. Finally, the NDVI image with a value between -1 and +1 was reclassified to define urban open green spaces. The classification data showed that open green areas have a nonhomogeneous distribution throughout the city, and the amount of green space per capita is approximately 200 m². According to the classification results of Northern Nicosia, a surface area of 475 km², the amount of green space is calculated as 16m², constituting 2.97%. Thus, it has been concluded that open green regions are denser in certain parts of the city than others and do not have a homogeneous distribution that can serve the whole town.

A Convolutional Neural Network Method for Rice Mapping Using Time-Series of Sentinel-1 and Sentinel-2 Imagery

Rice is one of the most essential and strategic food sources globally. Accordingly, policymakers and planners often consider a special place in the agricultural economy and economic development for this essential commodity. Typically, a sample survey is carried out through field observations and farmers’ consultations to estimate annual rice yield. Studies show that these methods lead to many errors and are time-consuming and costly. Satellite remote sensing imagery is widely used in agriculture to provide timely, high-resolution data and analytical capabilities. Earth observations with high spatial and temporal resolution have provided an excellent opportunity for monitoring and mapping crop fields. This study used the time series of dual-pol synthetic aperture radar (SAR) images of Sentinel-1 and multispectral Sentinel-2 images from Sentinel-1 and Sentinel-2 ESA’s Copernicus program to extract rice cultivation areas in Mazandaran province in Iran. A novel multi-channel streams deep feature extraction method was proposed to simultaneously take advantage of SAR and optical imagery. The proposed framework extracts deep features from the time series of NDVI and original SAR images by first and second streams. In contrast, the third stream integrates them into multi-levels (shallow to deep high-level features); it extracts deep features from the channel attention module (CAM), and group dilated convolution. The efficiency of the proposed method was assessed on approximately 129,000 in-situ samples and compared to other state-of-the-art methods. The results showed that combining NDVI time series and SAR data can significantly improve rice-type mapping. Moreover, the proposed methods had high efficiency compared with other methods, with more than 97% overall accuracy. The performance of rice-type mapping based on only time-series SAR images was better than only time-series NDVI datasets. Moreover, the classification performance of the proposed framework in mapping the Shirodi rice type was better than that of the Tarom type.

Comparative Analysis and Comprehensive Trade-Off of Four Spatiotemporal Fusion Models for NDVI Generation

It is still difficult to obtain high-resolution and fast-updated NDVI data, and spatiotemporal fusion is an effective means to solve this problem. The purpose of this study is to carry out the comparative analysis and comprehensive trade-off of spatiotemporal fusion models for NDVI generation and to provide references for scholars in this field. In this study, four spatiotemporal fusion models (STARFM, ESTARFM, FSDAF, and GF-SG) were selected to carry out NDVI image fusion in grassland, forest, and farmland test areas, and three indicators of root mean square error (RMSE), average difference (AD), and edge feature richness difference (EFRD) were used. A detailed evaluation and analysis of the fusion results and comprehensive trade-off were carried out. The results show that: (1) all four models can predict fine-resolution NDVI images well, but the phenomenon of over-smoothing generally exists, which is more serious in high-heterogeneity areas; (2) GF-SG performed well in the evaluation of the three indicators, with the highest comprehensive trade-off score (CTS) of 0.9658. Followed by ESTARFM (0.9050), FSDAF (0.8901), and STARFM (0.8789); (3) considering the comparative analysis and comprehensive trade-off results of the three test areas and the three indicators, among the four models, GF-SG has the best accuracy in generating NDVI images. GF-SG is capable of constructing NDVI time series data with high spatial and temporal resolution.

Unravelling net primary productivity dynamics under urbanization and climate change in the western Himalaya

Over the past few decades, climate change and urban expansion have strongly affected vegetation dynamics and overall carbon sink capacity of Himalayan ecosystem. However, the contribution of these two key factors on varying spatio-temporal scales in Himalayan landscapes still lacks in profound analyses. The present study takes Dharamsala and Pithoragarh urban landscapes as examples and uses the Enhanced Spatial and Temporal Adaptive Reflectance Fusion Model (ESTARFM) as the image fusion technique to generate highly resolved, both spatially (30 m) and temporally (monthly), NDVI images. These are used as inputs for Carnegie-Ames-Stanford Approach (CASA) model for Net Primary Productivity (NPP) estimation over the past two decades (2000–2020). During the study period, the NPP loss due to urbanization was 2065.43 kg C in Dharamsala and 401.07 kg C in Pithoragarh. Climate change, measured in terms of interseasonal changes in temperature, rainfall and solar radiation, had considerably contributed to the NPP variations of vegetated areas in both the study sites, however its contribution ratio was relatively lower than residual factors. There was a significant distribution gradient between how and where urban expansion and climate change influenced the NPP. Urban expansion impacts NPP more in plain areas corresponding to new urban land developments while climate change impacts NPP in high elevation mountainous regions dominated by Oak species. The observed declining trend of NPP under the current threat of urban expansion and climate change in the Himalaya highlights and invites the need of attributing importance to ecological issues. The findings of our study encompassing complex Himalayan landscapes could offer scientific perspectives for the management of ecological environment of highly fragile and vulnerable landscapes of the Himalaya.

Fusion of Different Image Sources for Improved Monitoring of Agricultural Plots.

In the Valencian Community, the applications of precision agriculture in multiannual woody crops with high added value (fruit trees, olive trees, almond trees, vineyards, etc.) are of priority interest. In these plots, canopies do not fully cover the soil and the planting frames are incompatible with the Resolution of Sentinel 2. The present work proposes a procedure for the fusion of images with different temporal and spatial resolutions and with different degrees of spectral quality. It uses images from the Sentinel 2 mission (low resolution, high spectral quality, high temporal resolution), orthophotos (high resolution, low temporal resolution) and images obtained with drones (very high spatial resolution, low temporal resolution). The procedure is applied to generate time series of synthetic RGI images (red, green, infrared) with the same high resolution of orthophotos and drone images, in which gray levels are reassigned from the combination of their own RGI bands and the values of the B3, B4 and B8 bands of Sentinel 2. Two practical examples of application are also described. The first shows the NDVI images that can be generated after the process of merging two RGI Sentinel 2 images obtained on two specific dates. It is observed how, after the merging, different NDVI values can be assigned to the soil and vegetation, which allows them to be distinguished (contrary to the original Sentinel 2 images). The second example shows how graphs can be generated to describe the evolution throughout the vegetative cycle of the estimated values of three spectral indices (NDVI, GNDVI, GCI) on a point in the image corresponding to soil and on another assigned to vegetation. The robustness of the proposed algorithm has been validated by using image similarity metrics.

Monitoring Land Vegetation from Geostationary Satellite Advanced Himawari Imager (AHI)

For many years, the Advanced Very High-Resolution Radiometer (AVHRR) and Moderate Resolution Imaging Spectroradiometer (MODIS) instruments have been widely used to monitor the condition of surface vegetation. Since the polar-orbiting satellite provides limited daily samples on surface, a completed spatial coverage of land vegetation is often relied on over multiple days of observations. In this study, observations from the Japanese geostationary satellite imager Advanced Himawari Imagers (AHI) are used to derive the surface vegetation index. The AHI reflectance at visible and near-infrared bands are first corrected to the surface reflectance by using the 6S radiative transfer model. The AHI surface reflectance from various viewing angles and solar geometry is further normalized to form an angular-independent reflectance by using a BRDF model. Finally, the surface vegetation index is calculated and synthesized from the daytime AHI data. It is found that the high-frequency AHI observations can significantly reduce the impact of clouds on compositing land NDVI and require a shorter time for a complete coverage of surface conditions. Also, a single NDVI image from AHI exhibits spatial distribution similar to that from 16 days of MODIS data.

Responses of net primary productivity to phenological dynamics based on a data fusion algorithm in the northern Qinghai-Tibet Plateau

Phenology is a key measure of how well an ecosystem functions and is pivotal in revealing vegetation productivity. Net Primary Productivity (NPP) and phenology have profound implications for the exploration of regional ecosystem processes. Several studies have investigated the effects of phenological indicators on NPP. However, these studies were carried out at a coarse resolution, making it difficult to obtain sufficient information on vegetation structure and dynamics; they also, potentially underestimate how productivity is influenced by phenology. This work uses fused NDVI images employing the ESTARFM model to determine how vegetation NPP reacts to phenology in the northern Qinghai-Tibet Plateau. The results showed that NPP was only 59.93 gC·m−2·year -1. The start of the growing season (SGS) was concentrated from March to May, whereas the end of the growing season (EGS) spanned from late September to early November. The length of the growing season (LGS) spanned 4–8 months during 2000–2020, NPP showed an increasing trend, SGS and EGS advanced, and LGS showed a prolonged trend. SGS and EGS were negatively correlated with NPP and seasonal temperature, however NPP was positively correlated with seasonal precipitation. LGS was positively correlated with NPP and interannual temperature, whereas it was negatively correlated with interannual precipitation. This study will aid in providing an accurate understanding of how NPP responds to phenological dynamics. Concurrently, collective understanding of terrestrial ecosystem responses to changes on a global scale will be advanced.

Analysis of time–space dynamic simulation on urban expansion in mountainous area based on the weight-CA model

As one of the central cities in the Wuling Mountains, Huaihua is regarded as the main body and core economic development area in the process of Wuling’s future new urbanization, exerting a pivotal influence on the five provincial border areas (Hunan, Hubei, Chongqing, Guizhou and Guangxi). But studies on urban expansion in this city are few, so this paper makes efforts to simulate its urban land expansion by constructing the Cellular Automata model (the Weight-CA model) taking into account differences in land types conversion. Firstly, the paper obtains the time-series Landsat spectral bands (Red, NIR) and NDVI images using the time–space fusion model; then the optimal combination of classification features is screened based on JBh distance and the object-oriented random forest algorithm is applied for land cover classification; the paper finally constructs the Weight-CA model for the simulation and validation of urban land expansion in Huaihua. Study results in the paper have shown that the Landsat spectral bands obtained with the ESTARFM have a relatively high correlation with real Landsat images. The land cover classification based on the optimal classification feature combination and the object-oriented random forest algorithm is high in accuracy, the total of which exceeding 84 % over 4 years. Meanwhile, based on the Weight-CA model, the simulation accuracy of urban land expansion in Huaihua in 2010 and 2015 reached 89.98 % and 89.60 % compared with the real values. It is operationally ready and practical and of strong practical significance to extract land cover types by Landsat and MODIS data fusion, and simulate urban land expansion using the Weight-CA method.

A spatiotemporal ensemble machine learning framework for generating land use/land cover time-series maps for Europe (2000-2019) based on LUCAS, CORINE and GLAD Landsat.

A spatiotemporal machine learning framework for automated prediction and analysis of long-term Land Use/Land Cover dynamics is presented. The framework includes: (1) harmonization and preprocessing of spatial and spatiotemporal input datasets (GLAD Landsat, NPP/VIIRS) including five million harmonized LUCAS and CORINE Land Cover-derived training samples, (2) model building based on spatial k-fold cross-validation and hyper-parameter optimization, (3) prediction of the most probable class, class probabilities and model variance of predicted probabilities per pixel, (4) LULC change analysis on time-series of produced maps. The spatiotemporal ensemble model consists of a random forest, gradient boosted tree classifier, and an artificial neural network, with a logistic regressor as meta-learner. The results show that the most important variables for mapping LULC in Europe are: seasonal aggregates of Landsat green and near-infrared bands, multiple Landsat-derived spectral indices, long-term surface water probability, and elevation. Spatial cross-validation of the model indicates consistent performance across multiple years with overall accuracy (a weighted F1-score) of 0.49, 0.63, and 0.83 when predicting 43 (level-3), 14 (level-2), and five classes (level-1). Additional experiments show that spatiotemporal models generalize better to unknown years, outperforming single-year models on known-year classification by 2.7% and unknown-year classification by 3.5%. Results of the accuracy assessment using 48,365 independent test samples shows 87% match with the validation points. Results of time-series analysis (time-series of LULC probabilities and NDVI images) suggest forest loss in large parts of Sweden, the Alps, and Scotland. Positive and negative trends in NDVI in general match the land degradation and land restoration classes, with “urbanization” showing the most negative NDVI trend. An advantage of using spatiotemporal ML is that the fitted model can be used to predict LULC in years that were not included in its training dataset, allowing generalization to past and future periods, e.g. to predict LULC for years prior to 2000 and beyond 2020. The generated LULC time-series data stack (ODSE-LULC), including the training points, is publicly available via the ODSE Viewer. Functions used to prepare data and run modeling are available via the eumap library for Python.